This research is designed to explore specific aspects of how individuals categorize spoken phonemes. The proposed experiments will measure the fine-grained aspects of phonetic categorization by distinguishing better and poorer exemplars of a phonetic category, and then testing the effects of different contextual factors on the category structure. The contextual factors to be tested include visual presentations of an articulating face that is dubbed with the auditory stimuli, acoustic cues in neighboring phonetic segments, and phonotactic constraints; subjects will be asked to rate synthetic speech stimuli that differ along specific dimensions on how well they represent a particular phonetic category. Statistical analyses will determine whether the contextual factors alter which members of a phonetic category are considered the best exemplars of the category. The general aim of the research is to examine the contributions to speech perception of different types of information sources, using techniques capable of uncovering fine-grained aspects of phonetic structure. These techniques are expected to contribute new knowledge relevant to debates regarding the different possible roles that these information sources might have in perception. Thus, this research will contribute to the development of general theories of speech perception. The information gained from this research will also have important implications for the treatment of hearing impairment, by specifying properties that are less or more crucial for the ability to distinguish phonemes.